tante

https://tldr.nettime.org/@tante

Sociotechnologist, writer and speaker working on tech and its social impact. #Communist. #Feminist. #Antifascist. #Luddite. Email: tante@tante.cc | License CC BY-SA-4.0 #noAI

"Ein-Mann-Gegenkultur" (SPIEGEL)

website: https://tante.cc
pronouns: he/him
contact: https://tante.cc/contact/

Public Key

npub1j8vya2y336yh2fkxqmepgf704l3lxd5rtqml7g0fndx7umrlc22qvkld5q

NIP-05 Address

tante_at_tldr.nettime.org@momostr.pink

Profile Code

nprofile1qqsfrkzw4zgcazt4ymrqdus5yl86lclnx6p4sdlly85ekn0wd3lu99qpz3mhxue69uhhyetvv9ujuerpd46hxtnfduqs6amnwvaz7tmwdaejumr0dszkaw4n

Publishing to

relay.damus.io

nos.lol

Author Public Key

npub1j8vya2y336yh2fkxqmepgf704l3lxd5rtqml7g0fndx7umrlc22qvkld5q

Show more details

Last Notes

2026-08-21 10:05:49 UTC

Also made some of @npub1l77…cjyv's cards. For my sanity's sake.
https://tldr.nettime.org/system_nettime/media_attachments/files/117/132/930/142/461/528/original/f0839ef389ffbb7f.jpg

2026-08-18 21:24:50 UTC

- reply

oh, who showed their asses now?

2026-08-17 06:33:07 UTC

"Claude watermarks your code now."
Dude. It's not yours, you didn't write it.

2026-07-22 08:48:57 UTC

No, OpenAI's new magic models did not autonomously hack Huggingface.
Per OpenAI's PR blog post (https://openai.com/index/hugging-face-model-evaluation-security-incident/):
"This incident occurred during an internal evaluation which prompts models to pursue advanced exploitation using complex attack paths, in an effort to quantify their cyber capabilities. We estimate maximal cyber capabilities by running this evaluation without production classifiers used to prevent models from pursuing high-risk cyber activity."
So they prompted their model running without guardrails to hack some shit. OpenAI told it to do that, the model didn't do shit autonomously (no LLM ever does anything autonomously, it's always prompted).
The whole story is PR. We know that from Anthopic's Mythos: "Look we have this super secret new model and it's so powerful. We are scared ourselves. And you will soon be able to rent it!"

2026-07-22 07:30:13 UTC

RE: https://tldr.nettime.org/@tante/116880003584050912
It passed.
https://codeberg.org/Codeberg/org/pulls/1253#issuecomment-19820434
Codeberg bans vibe coded slop.
(EDIT: No, I do not want to debate the percentages of LLM code allowed or "how will people enforce this, I can still vibe code and not tell". This is a statement about intent by a community. It will be a moderation challenge, especially with so many people who have made "coding with an LLM" their whole personality but it's doable. But don't waste your and my time with these boring, juvenile arguments)
#note1n9n…af8y

2026-07-07 18:02:00 UTC

If you are a member of Codeberg e.V. please take the time to participate in the poll that was just sent out about banning vibe-coded projects on Codeberg.
Please agree to the proposal. Slop can live on GitHub.

2026-07-03 16:32:49 UTC

RE: https://mastodon.social/@vicgrinberg/116856970122743941
While the general statement is true, the line underneath undercuts the statement a bit. "I use these tools as well all the time". Okay but if you know that they undercut your ability to think ... why? What the fuck are we doing?
#note19qz…hktz

2026-06-30 13:26:35 UTC

The "AI Compass" is cute. But I am disappointed not to get put into the "Luddite" group. And I personally wouldn't have made an artist that went all in on NFTs the "patron saint" of "The Conscientious Objector"
https://bambamramfan.github.io/ai-compass/
https://tldr.nettime.org/system_nettime/media_attachments/files/116/839/284/353/638/466/original/e03b3e83c9abcdff.png

2026-06-10 07:40:47 UTC

A court in Munich declared that Google is liable for their "AI summaries" and all its hallucinations. This is an important step to bring "AI" slop in line with all other products on the market: "AI" products are basically the only ones where a provider can just deliver unchecked garbage and put all the liability on the consumer. I hope to see aggressive change here.
https://the-decoder.com/landmark-german-ruling-declares-googles-ai-overviews-are-googles-own-words-and-makes-it-liable-for-false-answers/

2026-05-20 08:25:56 UTC

RE: https://tldr.nettime.org/@tante/116605858023186072
I wonder if we'll see a new term for "the web" that's somewhat like "the dark web". Something that marks the open web as unclean, unruly, dangerous, bad and only marks the LLM filtered version as desirable.
#note1c7e…x0zw

2026-05-15 22:01:33 UTC

We sometimes fall into the trap of distinguishing art from "AI" slop based on the effort and work the artist put in.
But of course that's not fully it: You can spend a lot of time iterating with an LLM and it's probably still not art.

2026-05-07 11:44:57 UTC

"There are no more juniors. There was a funeral for their passing in 2024. Nobody came. The machine does what they do now, but cheaper. Of course, juniors weren't valuable for what they produced, they were valuable for who they would become: the senior engineer who knows where the bodies are buried. We optimized for output, and abolished apprenticeship. A few years from now, we'll wonder where all the seniors are. We shot them. Nobody will remember."
https://www.stvn.sh/writing/programming-still-sucks-fqffhyp

2026-04-29 12:58:40 UTC

The Open Source community never learned that big tech is not trustworthy.
Maybe this time?
(based on https://de.wikipedia.org/wiki/Datei:Zuckerberg_-_dumb_fucks.png)
https://tldr.nettime.org/system_nettime/media_attachments/files/116/488/111/320/619/976/original/93f42305253181ca.png

2026-04-27 10:56:28 UTC

RE: https://masto.ai/@phoronix/116476120066302561
Okay gotta migrate off to stock Debian then I guess.
Microsoft kills its OS with all the copilot garbage users do not want and Canonical has to go: "Yeah, that's where we need to go."
Strategic genius.
#note1xed…zk88

2026-04-24 12:43:54 UTC

A lot of the time when people argue that we should look at the opportunities of "AI" instead of always looking at the damages, that "AI" just won't go away I mentally replace "AI" with "child labor" and it really shows how people think about the world.

2026-04-21 21:14:13 UTC

In which I argue that "AI" as we understand it today is a structurally fascist artifact.
https://tante.cc/2026/04/21/ai-as-a-fascist-artifact/

2026-04-21 06:07:29 UTC

RE: https://pdx.social/@dadbod/116437799506135617
This sort of repairability should be legally required.
#note1ldn…4wgf

2026-04-20 17:30:40 UTC

The "agentic" in "agentic AI" refers to the agency the "AI" providers want to take from you.

2026-04-20 07:24:19 UTC

I have been warning about the term "digital sovereignty" and how it is right-wing coded and probably can't be salvaged for non-right-wing politics.
The German fascist party AfD now created a European foundation to push their fascist politics further. The name: "Sovereignty Foundation".
Even though tactically it might feel like it makes sense to use the term to get funding, you are integrating right wing politics into your thinking and speaking.

2026-04-16 14:47:34 UTC

RE: https://tldr.nettime.org/@tante/116414897025610603
"Tech" wasn't /taken/ from us. We just never were willing to realize that it did not include us except as powerless users and customers.
Love for "Tech" is often more akin to Stockholm syndrome.
#note1mj9…r2zp

2026-04-16 14:39:17 UTC

I do not want to "take back tech".
I want something new to grow from the needs, utopias and desires that a community focused on a convivial life develops.
"Tech" as a social structure might not be salvageable.

2026-04-08 13:44:39 UTC

"On the acceptance of GenAI"
https://smallsheds.garden/blog/2026/on-the-acceptance-of-genai/
https://tldr.nettime.org/system_nettime/media_attachments/files/116/369/383/660/367/551/original/0c4cfd13cf91f6be.png

2026-04-08 13:18:24 UTC

"But X is useful" is not the good defense you think it is.
Who is it useful to, for what purposes and who bears the costs/harms?
Child labor is useful (to people wanting to squeeze out some money from desperate people). We still think that it should not exist.

2026-04-08 12:58:15 UTC

An "AI" company saying "we have this super magic new model that is so powerful that we cannot release it" is just marketing. It's the old OpenAI scam.
Anthropic needs to keep the hype going to ride it to the IPO. That's all this is.

2026-04-08 08:40:04 UTC

"How to make a fortune with AI"
(Original title: Make a fortune with AI)
https://designthinkingcomic.substack.com/p/make-a-fortune-with-ai

2026-04-08 07:59:00 UTC

Oh, Anthropic is Oneshotting people again with detected security flaws?

2026-04-02 17:09:22 UTC

- reply

they do not.
I recently set up Ghost for a test and it's nice UI wise and all. But the setup is so complicated and fragile ... I've been running services on Linux boxes for more than 2 decades now and I failed twice to get it to work.

2026-04-02 16:45:55 UTC

- reply

also: The sales pitch for WordPress is buckets of plugins and it runs everywhere.
Now they have something without those plugins that is massively harder to run. What are they even doing?

2026-04-02 16:37:09 UTC

- reply

it is not a fork, is it? Just another probably vibe coded reimplementation to strip a GPL from popular code.

2026-04-01 15:00:05 UTC

"The experience of workslop directly erodes social connection, erodes trust in one another and in the end erodes solidarity. Because why would you stand with a person who does not do their job and offloads their work on you??"
(Original title: Dissolving the social)
https://tante.cc/2026/04/01/dissolving-the-social/

2026-03-31 21:14:04 UTC

Anthropic's Claude Code's full source code leaked. Claude is seen by many to be the best coding LLM on the market with Anthropic proudly stating that Claude Code itself is mostly written by the LLM.
Now this sounds good as long as nobody can see the code which is quite the trash fire. Detecting "code sentiment" via regular expressions, variable and functions names containing prompt parts trying to influence the bot, a completely intransparent mess of a control flow that makes actual maintenance and debugging functionally impossible and the prompts ... of the prompts. All the begging and pleading to the chatbot not to do this or not to do that or please to do this.
It is fascinating but it is as far away from actual engineering as drunkenly pissing your name in the snow. Dunno what you call the people prompting software at Anthropic but "engineer" is not it.
Now it is fun to look at the currently hyped product striped bare and showing its pathetic quality but that is the future of software if we let those companies continue to undermine every good practice software engineering has tried establishing.
The software we have to use will be bad, insecure, unmaintainable, expensive with nobody having the skills or resources to build something better. As I wrote a few months ago: LLM based software production is equivalent to saying that fast fashion should be the only way to produce clothing. A tragic degeneration of the quality of the artefacts we rely on build for maximum profit on the backs of people in countries from the global majority.

2026-03-31 17:44:56 UTC

RE: https://neuromatch.social/@jonny/116324988415693534
Had the same impression. It's just a bunch of nerds praying to their machine god to do things right. Please. Pleasepleaselplease.
These are not serious people. And they are not engineers for sure.
#note10sz…yj93

2026-03-25 15:00:46 UTC

I personally consider "I asked ChatGPT to generate a response to you" not witty but a form of an insult. Don't do that please. If that is how you want to talk to people at least don't tell me. It's offensive.

2026-03-24 20:14:43 UTC

OpenAI kills their Sora Video generator.
Which isn't that surprising TBH:
- it is massively expensive to run
- the results were not that usable
- it was a lawsuit generator
So in order to keep the numbers looking better till their IPO at the end of this year they are starting to cut shit. Just an indicator that things are not looking good cashflow-wise.

2026-03-23 13:09:35 UTC

RE: https://tldr.nettime.org/@tante/116278629466465183
Spending 3 to 5 USD for every USD you make is not exactly what I'd consider good business.
#note1l6l…agkp

2026-03-23 13:04:39 UTC

"Microsoft 365 has more than 450 million paid commercial seats. After roughly two years on the market, Copilot has converted approximately 15 million of them into paying users. That's a 3.3% conversion rate, at $30 per user per month, generating roughly $5.4 billion in annual revenue. That's less than what Microsoft spent on infrastructure in a single quarter (3)."
None of the numbers make economic sense.
https://finance.yahoo.com/markets/stocks/articles/nadella-paid-650m-recruit-ai-170000361.html?guccounter=1

2026-03-22 14:30:14 UTC

It was my plan to read some Ursula K. Le Guin and having gone through half of "The Dispossessed" I am so angry for it taking me so long to do it.
Holy fuck.

2026-03-20 12:20:06 UTC

Companies are realizing that pushing people to using "AI" is expensive (even at the subsidized pricing going on right now)
(Original title: Companies go full AI — then the bill comes due)
https://pivot-to-ai.com/2026/03/19/companies-go-full-ai-then-the-bill-comes-due/

2026-03-20 11:04:39 UTC

Marc Andreessen claiming that he "doesn't do introspection" is just another building block in the attempt to shift the definition of "person" in a way to allow "AI"s to gain rights.

2026-03-19 17:54:59 UTC

RE: https://wandering.shop/@xgranade/116257004378454654
This is garbage. `uv` is the best tool for the job in the Python ecosystem but it being developed and maintained by a startup (with no path to sustainable funding) this was basically inevitable.
People underestimate the bad position open source is in currently and how much of that space is annexed
#note1nsk…v5fd

2026-03-18 12:40:18 UTC

Audre Lorde's "The master's tools will never dismantle the master's house." is not just a statement about a tool being tainted by its origin. It's about what kind of tool a "master" would create: Whips. Chains. Violence.
That's the meaning: You cannot just take tools whose purpose and politics is dominance and violence and "make them liberatory". This goes deeper than "just" embedded politics or lofty talks about ethics, it comes down to what kind of relations you believe do and should and must not structure the world.

2026-03-16 18:57:50 UTC

RE: https://mastodon.social/@anon_opin/116240371365473711
I have not. If I have to pay _and watch ads_ in my own home, fuck you and you do not get my money anymore.
I also don't watch stuff with ads anymore. It's unbearable after you detoxed.
#note1xdp…w86q

2026-03-16 18:56:17 UTC

RE: https://tldr.nettime.org/@tante/116239395175554368
So yeah. I am kinda looking for a new job that allows me to make our digital world meaningfully better. If you have any leads or thoughts, ping me.
#note1w6r…fwuj

2026-03-16 14:46:51 UTC

Day of rejections.
Didn't get a job and won't be speaking at Re:publica about decomputing.
https://tante.cc/2026/01/19/republica-2026-talk-proposals/
So, sadly no #rp26 for me

2026-03-12 21:27:16 UTC

One of the weirder things in "AI" is the whole "skills" thing. It sounds as if those were somehow little programs or tools integrated into your agents but they are just markdown files with strings that you pull into your prompt to hope the slop machine does better. Crazy shit.

2026-03-12 18:32:07 UTC

RE: https://social.coop/@cwebber/116217477822586442
The whole "AI" industry is just based on the hope that they can deskill people fast enough that they'll have to rent back cognitive support systems. Forever.
It's like Uber. Just that they don't try to break existing transportation infrastructures but your brain.
#note1qcn…g8v4

2026-03-11 18:51:24 UTC

- reply

Which was really fucked up to see: These folks actually want to protect their organizations from burning a lot of resources on bullshit instead of fixing actual problems that help the workers _and_ the organization. And they have to actively fight management who got their brains ruined on linkedin.

2026-03-11 18:48:35 UTC

- reply

But it was super fun to lead them through a "this is how you can force reasonable evaluation on 'AI' projects which kills most of them" framework and see how they felt empowered and able to actually do their job again.

2026-03-11 18:45:20 UTC

Spend the day talking to workers council members about "AI". And it's kinda wild hearing their stories from the wild: Management is 100% in fantasy "AI" can do everything land and makes huge plans for how to use "AI" to cut workers when real projects that supposedly can do 50% of a specific task end up being able to do 8%. And they still go live. It's fucking bonkers. CEO's are really not okay.

2026-03-11 09:02:48 UTC

"Silicon Valley is buzzing about this new idea: AI compute as compensation"
Of course they are. "AI" people are Blockchain people and they'd love to force people into money that _they_ control
https://www.businessinsider.com/ai-compute-compensation-software-engineers-greg-brockman-2026-3

2026-03-10 20:52:30 UTC

RE: https://tldr.nettime.org/@tante/116206844690350737
So yeah. Kinda looking for a new role in something mission driven where I can actually contribute something meaningful.
#note1dcr…6s0k

2026-03-10 20:48:50 UTC

At work the development team keeps getting more and more into "agents" and I am increasingly dreading those meetings because I feel it affecting my relationship to people.

2026-03-10 20:03:05 UTC

"You go to sleep, the agent does the science and when you wake up, you have the results."
That's not how any of this works. You're not "doing research with help" you have something generated that looks like a research report. Without the research happening. How many different ways can "AI" bros find to express "I AM TOTALLY MISSING THE POINT"?

2026-03-09 14:56:57 UTC

New names:
Anthropic's model are now called TAGS (targeting a girls school) OpenAI's are now going by NUPTOP ("nudging people to psychosis")

2026-03-09 13:42:14 UTC

"AI" skills are basically just documentation and it says something about how documentation only gets resources and respect if it is framed as "technology". Which has gender bias all over it.

2026-03-09 13:14:30 UTC

RE: https://mastodon.green/@gerrymcgovern/116199381812682900
That is an interesting thought. I myself am in that kind of age range and this theory does explain a few things I am seeing with some peers.
#note15g4…h2jf

2026-03-09 13:00:28 UTC

A few unsorted thoughts about how the "Claude" models differ from what for example OpenAI release. Basically: One is a religious text, the other tries to be more of a commercial product
https://tante.cc/2026/03/09/claude-magic/

2026-03-05 10:02:07 UTC

Okay, my "Democracy instead of just “Open Source”" talk proposal for #rp26 was rejected. Maybe I'll find another home for it.
https://tante.cc/2026/01/19/republica-2026-talk-proposals/

2026-03-05 08:25:01 UTC

The "Pro-Human AI Declaration" that was released is kinda bland and not very consistent or meaningful (https://tante.cc/2026/03/05/nothing-to-declare/) but the core takes now is to hold the people in your circle who thought that signing this together with Steve Bannon and other racists was a good idea accountable.

2026-03-04 14:28:40 UTC

Yesterday I read an interview with the singer in a punk band who uses GenAI to help with writing his texts.
Punk. The whole point of punk was that if you know how to play 3 cords badly and have something to say just fucking go for it.

2026-03-04 12:55:08 UTC

RE: https://tldr.nettime.org/@tante/116170621153319970
Kept thinking about it so I tried to add to it in a quick article. It's about software and love. And Open source.
https://tante.cc/2026/03/04/artisanal-care/
#note1t0w…zqyf

2026-03-04 11:16:43 UTC

- reply

Here's the thing: I believe that you deserve to have access to high quality products and services. You deserve to use products and services that are safe, secure, well-designed and not destroying the ecological, informational or social environment.

2026-03-04 11:14:11 UTC

The question is not whether you can create software using LLMs - you can - most software is just boring CRUD shit.
But you do pay a hefty price: In lowering quality (security issues, less maintainable), in skill decay in the people "guiding" the stochastic parrots, etc.
It's not "can 'AI's create software" but "are we willing to accept worse software running more and more of our lives?"

2026-03-03 12:43:18 UTC

As a human being you have the right to be able to understand the world. That is why education is so important.
Any technology that hinders your understanding (even if it seemingly gives you capabilities) is deeply in- and anti-humane.

2026-03-03 12:40:52 UTC

RE: https://tldr.nettime.org/@tante/116165239429495379
I think that is actually the biggest problem with those systems.
It's not losing some random skill that we have made superflous, it's about taking skills that are the foundation for your agency in this world and your ability to understand it.
It is a fundamental human right that is being taken from you for the sake of efficiency and convenience.
#note160e…9z7x

2026-03-03 12:28:04 UTC

RE: https://hachyderm.io/@thomasfuchs/116165215922836572
Their makers also do not want you to be able to understand. It's about turning the means of understanding and sensemaking into externalized products to rent.

2026-03-03 11:16:26 UTC

The answer to problems in sociotechnical systems is not always "more software", "better software" or "open source software".
A lot of the time the best solution is _less software_ and sometimes even _no software_.

2026-03-02 11:00:05 UTC

"we need to open our eyes to the fact that the current industry many of us work in, not only doesn’t care about their workers, it actively resents them. In their eyes, we have gone from being the people who made things possible, to an unnecessary burden on the bottom line.
They hate that we charge money for our labor, and see that money as something we are stealing from their pockets."
(Original title: How to grow strawberries)
https://buttondown.com/monteiro/archive/how-to-grow-strawberries/

2026-03-02 08:35:32 UTC

- reply

on the other hand: I test these systems and have colleagues using them but I don't use any LLMs for any of my work or activities.

2026-03-02 08:34:51 UTC

- reply

I found its rate of very clear falsehood generation quite bad and it works even worse than other models in the contexts people use those systems (think coding assistant, rephrasing text, etc)

2026-03-02 08:28:50 UTC

- reply

performance ain't great.

2026-02-27 09:50:24 UTC

Given how they can only be produced by exploitation of workers from the global majority no system including any of the big LLMs in their production can ever be called "fair".
"Fair LLMs" of the size required to do the tasks people want LLMs to do (badly) do not exist.

2026-02-26 10:26:41 UTC

- reply

This is not criticism of the project or the people doing that (soul-crushing) work. It's criticism of the funding structure and metrics applied to research and other public funding contexts.

2026-02-26 10:24:37 UTC

- reply

"Our post about a presentation got 37 likes on LinkedIn".
Oh, we're basically influencers now.

2026-02-26 10:21:55 UTC

At work we are part of a EU funded research project. These days it's actually expected to track your social media presence for "impact". So I have to listen to a 10 minute report by someone counting the amounts of likes some meaningless post got on instagram. This is a good investment of everyone's time and money.

2026-02-25 20:17:23 UTC

"Kyle Fish, who leads model welfare research at Anthropic"
Journalism is hard. How can you write that sentence without wanting to shoot yourself in the head 4 or 7 times.
Yeah sure. Your bucket of statistics needs a spa day. Dipshits.

2026-02-25 11:35:05 UTC

RE: https://mstdn.social/@jukkan/116130658933405707
It's not just people at Meta running irresponsible vibe coded software. Microsoft ain't any better
#note1ey6…pq7u

2026-02-25 07:20:04 UTC

- reply

Firefox's translation engine is a small LLM that runs on your device. Since it uses the website your browser already downloaded as input it is not affected by iocaine

2026-02-24 11:54:41 UTC

The Times New Resistance font is a work of art.
https://www.abbyhaddican.com/times-new-resistance

2026-02-24 09:59:19 UTC

RE: https://tldr.nettime.org/@tante/116124997771919604
Everyone loves not having to have a license to develop and release code but the argument for stronger controls on who gets to do surgery on the core infrastructures of our digital society is getting stronger by the day.
#note18h0…kyqu

2026-02-24 09:55:27 UTC

Basically: If you run OpenClaw connected to any meaningful system you are not fit to design, program or run any kind of software. That disregard for security and quality should leave a black mark on you for many years.

2026-02-24 09:54:05 UTC

RE: https://flipboard.com/@404media/404-media-qvt3vv94z/-/a-WUaas_e8S32MEVyncTs4mA%3Aa%3A4082434389-%2F0
This is a great illustration about the state of the art in "AI": People with access to inconceivable resources running massively expensive, dangerous code on anyone's data (sometimes even their own) with no regards for basic security or quality standards and practices.
The discipline of software engineering has never been great at establishing and following those kinds of practices but "AI" has thrown us back decades. And we'll suffer the drop in software quality and security for a long time.
#note17n6…prwd

2026-02-24 06:40:41 UTC

"Briggs and his Goldman Sachs colleagues recently said that investment spending on AI made “basically zero” difference in U.S. economic growth last year."
As Robert Solow was saying ...
https://www.washingtonpost.com/technology/2026/02/23/ai-economic-growth-gdp-mirage/

2026-02-23 14:08:51 UTC

Worth repeating: Claims of "AI" democratizing anything (coding, creative endeavors, etc) are always BS. The thing locking out people from doing that is not having the time/resources.
You want to democratize coding/art/creativity? Give people paid time off to do it and access to infrastructure. Easy.

2026-02-23 13:47:28 UTC

At least 70% of things I read about people doing with "AI" assistants like "Open Claw" are basically things people have been doing with deterministic, secure tools like Home Assistant for years.
Oh you can tell your bot to dim the lights via WhatsApp chat? Welcome to 2015!

2026-02-23 10:02:31 UTC

Ich war zu Gast bei Breitband vom DLF Kultur und wir sprachen (auch) über "Digitale Souveränität" und Open Source.
https://www.deutschlandfunkkultur.de/jugendschutz-in-social-media-die-plaene-der-plattform-discord-100.html
TL;DL: Ich halte das Thema "Digitale Souveränität" viel weniger für ein Shoppingthema (wo kaufe ich meine Software) denn eine Frage der Fähigkeiten einer Organisation, IT Systeme entlang ihrer Ziele und Bedüfrnisse aktiv zu gestalten. Open Source ist dazu eher notwendig denn hinreichend.

2026-02-23 09:24:32 UTC

- reply

Das (vorgeschobene) Versprechen von Automatisierung ist ja die Sicherung und Standardisierung von Qualität: Es kommt verlässlich ein gewisser Qualitätsstandard heraus. Das ist eine der zentralen Legitimationen von Automatisierung (neben der echten: Arbeitnehmermacht reduzieren und Kosten drücken).
Wenn man das nimmt, dann gehts halt wirklich nur noch ums Kosten drücken. Das ist in einem Bereich wie Jura, wo es um existenzielle Bedürfnisse und Rechte von Menschen geht, ziemlich fatal. Wäre jetzt nicht mein Move da so mit "naja, is ja eh so" drüber wegzugehen.

2026-02-23 09:14:32 UTC

RE: https://mastodon.social/@AwetTesfaiesus/116119030408309460
"Menschen betrügen manchmal, deshalb darf man eine Maschine, die ziemlich verlässlich falsche Informationen verbreitet nicht deshalb kritisieren" ist ein ziemlicher Hot Take für eine *checks notes* Bundestagsabgeordnete.
Selbe Logik: "Menschen fahren manchmal zu schnell, daher dürfen Selbstfahrende Autos so schnell fahren, wie sie wollen." oder "Menschen machen manchmal den Grill an und erzeigen CO2, daher braucht eine automatisierte Produktionsstraße sich nicht an CO2 Limits zu halten".
Das ist eine Kapitulation. Die Aufgabe politischen Gestaltungswillens weil man an Technikdeterminismus glaubt.
Nicht ideal für die Obfrau des Ausschusses für u.a. Medien.
#note1tfu…3gj8

2026-02-22 18:38:39 UTC

Many of the people I do have online conversations with speak English as their first language. I kinda get by but you native speakers really underestimate how hard it is to express oneself in a foreign language. Concepts work differently, metaphors don't really translate, references you have used for decades don't make sense.

2026-02-22 12:34:10 UTC

Dear media writing about tech: Sam Altman shitting on "data centers in space" is not the story. Nobody thought that that narrative makes any sense.
You failed as reporters when you pretended it did just cause rich dude claimed it would be the future.

2026-02-21 12:20:05 UTC

"Capable LLMs require a logic of dominance and of disregarding consent of the people producing the artifacts that are the raw material for the system. LLMs are based on extraction, exploitation and subjugation. Their politics is violence. How does one “liberate” that? What’s the case for open source violence?"
(Original title: Acting ethically in an imperfect world)
https://tante.cc/2026/02/20/acting-ethical-in-an-imperfect-world/

2026-02-20 21:39:54 UTC

RE: https://tldr.nettime.org/@tante/116102639569199871
Follow up to the article from this morning. In which I argue how @npub1fdr…lvhs was right in his article: https://tante.cc/2026/02/20/on-alliances/
#note1vxm…wyns

2026-02-20 13:37:32 UTC

- reply

on this one for example I fully agree with Cory. This is not him having a genAI system write or anything like that.

2026-02-20 11:08:06 UTC

Yesterday Cory Doctorow argued that refusal to use LLMs was mere "neoliberal purity culture". I think his argument is a strawman, doesn't align with his own actions and delegitimizes important political actions we need to make in order to build a better cyberphysical world.
https://tante.cc/2026/02/20/acting-ethical-in-an-imperfect-world/

2026-02-19 23:53:14 UTC

- reply

Let's see if I write this up properly tomorrow morning and put it on the blog in order to have a bunch of fanboys tear me to shreds.

2026-02-19 23:50:06 UTC

- reply

Dammit, now I kinda got into it.

2026-02-19 23:49:24 UTC

- reply

Finally: It's easy to say "but crawling and taking everything is fine, it's what a search engine does". But that requires so much abstraction that it almost becomes a lie: A search engine processes data in order o generate pointers to it, the "index". LLMs do explicitly not do that but annex, conquer, take while cutting any connection the original author/source has to their audience.
The technology might be the same, the systemic effect, the politics are radically different.

2026-02-19 23:46:34 UTC

- reply

This whole argument says that anything you ever do is okay, as long as something "free and open" comes out of it. But the free and open torment nexus is a torment nexus and needs to be destroyed.
Also technology does not need to be "liberated" (and often cannot, artifacts do have politics). If anything needs liberating it's people. And I don't see how LLMs (even the open weights ones) are doing that.

2026-02-19 23:43:56 UTC

Not gonna get into it right now (gotta go to bed) but labeling criticism of LLM based "AI" as "purity culture" and that one can just legitmize using any and all tech if one just somehow creates "free and open" versions of it is not a good take. Really not. Refusing to use LLMs on ethical grounds is also not a claim that problems are solved "by shopping carefully". That's a lot of straw men just to legitimize using an LLM to do spellcheck. Maybe just do it and don't try to throw people under the bus who aim for something better.
[Edit: removed a snarky comment that didn't need to be there]

2026-02-19 16:01:56 UTC

I am a luddite and to me there is a lot of joy in technology. In technology that we deploy for all our wellbeing. Solar panels and vaccines, high-speed rail and wikipedia.
But they joy stems from the feeling of being able to be part of humanity in an embedded, meaningful way. Not from buying shit.